Tag Archives: tyler cowen

Polarized electorates and McDonald’s salad

Tyler Cowen takes issue with Cass Sunstein’s new book, Going to Extremes, which postulates that American politics is stuck in a polarized posture, with the two antagonistic parties having little incentive or inclination to reach compromise. Cowen lists some reasons to think politics in the U.S. are not getting more polarized, including

6. Obama goes out of his way to adopt a non-polarizing style (no matter what you think of his policies) and it brings him considerable popularity.  That suggests a demand for non-polarization, or at least the perception thereof.  In many countries politicians have an incentive to straddle the median and bring outlying groups closer to the center, for purposes of governance and re-election.

Now let me pretend to be a behavioral economist: the phrase  “or at least the perception thereof” is doing a lot of work here!  What we know is that there’s a demand for the ability to believe about oneself that one is a post-partisan and unpolarized type.  But that’s not to say there’s a demand for actual unpolarized politics.  People tell McDonald’s market researchers they’d like to see more salad on the menu; they’re more likely to go to McDonald’s when there’s salad on the menu; but they don’t eat the salad.  In fact, the presence of the salad option appears to make customers more likely to order a less nutritious meal.  Apparently they feel they’ve done enough for their health by being in the same room with the salad.

If Obama is the salad and nutty partisan views are the fries, Cowen’s argument might actually work in Sunstein’s favor:  we are all in the room with the salad, and thus might feel freer to indulge our cherished political idiosyncracies.

In a different direction, one might set aside the question of what market and structural forces might in theory push politics to be more or less polarized, and ask instead whether one can quantify what’s actually happening.  Cowen says that public opinion polls about polarization point in both directions.  I like the work of McCarty, Poole, and Rosenthal on this question;  they have a nice way of modeling the U.S. Congress as a subset of R^2, in which setting you can see the two parties getting further apart and more ideologically disciplined over time.   Here’s what I wrote about their techniques in Slate in 2001.    On their view, at least, it’s not in question that political polarization has been increasing since the 1960s.

Tagged , , , , , , , ,
%d bloggers like this: